When semiconductor chips having a multiplicity of contacts, so-called inputs and outputs (“IOs”), are linked to a substrate, a so-called “fan-out” or “fan-in” is required in order to link all inputs and outputs, respectively, since plated-through holes, so-called vias, in substrates cannot be realized with anything like the same density as the inputs and outputs on a semiconductor chip. In the case of “fan-out” or “fan-in”, conductor track traces are in this case spread apart or drawn together, which is also designated as “redistribution wiring”.
Each “fan-in” or “fan-out” usually requires a so-called solder stop, since solder during reflow soldering would otherwise flow away over the conductor track in an undefined manner. In the case of flip-chip bonding, this would lead to what is referred to in the jargon as collapsing of the solder bumps. By way of example, a layer composed of a poorly wettable polymer over which the solder does not flow away is used as a solder stop. The polymer layer has to be photopatterned on account of the required accuracy. Usually, with the use of a solder stop, both so-called SMD arrangements (SMD: “solder mask defined”) and so-called NSMD arrangements (“non-solder mask defined”) are used, wherein soldering regions are predefined by a solder mask in the case of an SMD arrangement, whereas the soldering regions are determined by the conductor track traces themselves in the case of the NSMD arrangement.